During visual system development, retinal afferents must not only find their correct topographic location within the target, but there must also be a correct numerical match between afferents and target, even though both are initially overproduced. This control process is critical in establishing visual receptive field properties and perceptual acuity. Studies outlined in this proposal will examine the control process, using the hamsters retinotectal system as an experimental model. Partial ablation of the superior colliculus in neonatal hamsters creates a population mismatch between retinal axons and the remaining target tissue, which serves to compress the visual field map but which leaves largely unaffected the receptive field properties of individual collicular neurons. The PI proposes that the conservation of receptive field properties is the result of activity-dependent preservation of convergence ratios between retinal axons and collicular neurons. Much of this conservation mechanism is not visually driven since the process occurs prior to eye opening. NMDA receptors are a potential component of the conservation mechanism. Through its role in coincidence detection the NMDA receptor may provide collicular neurons with the ability to recognize and select for competing retinal inputs with temporal activity patterns that fall within a specific time window (which translates into a window of visual space). The model predicts that the window varies in width between central and peripheral representations of the visual field. Such a mechanism would enable the preservation of visual response properties without interfering with the formation of a compressed retinal map. The goal of the proposed studies is to test the model by examining retinal map formation and the development of visual receptive field properties under conditions where activity of retinotectal axons is blocked, patterned visual input is disrupted and NMDA receptor function is antagonized. Potential variations in temporal requirements for activity-dependent cooperation with eccentricity will be examined by varying the interstimulus interval between two visual inputs and examining the effect of this variation on synaptic potentiation in neurons located throughout the superior colliculus. Because the proposed studies address questions not only of map formation but also of response property construction they uniquely begin to look for developmental control of complex functional capacities across visual areas of the CNS. These control mechanisms may protect against perinatal damage by allowing the visual system to compensate for damage to a target cell population without compromising visual function.